Presentation is loading. Please wait.

Presentation is loading. Please wait.

Delay Tolerant Data Collections Bathymetry in Google Earth dtnrg 2012 Kurt Schwehr Google / UNH CCOM

Similar presentations


Presentation on theme: "Delay Tolerant Data Collections Bathymetry in Google Earth dtnrg 2012 Kurt Schwehr Google / UNH CCOM"— Presentation transcript:

1 Delay Tolerant Data Collections Bathymetry in Google Earth dtnrg 2012 Kurt Schwehr Google / UNH CCOM http://schrwehr.org

2

3

4

5 Cosco Busan and the tug Revolution Cosco Busan, Nov 2007

6 As a community, are we ready for another difficult incident?

7

8

9 DTN for Ocean going instruments is not new

10 Water level - Tide Aware/Time dependent path

11

12 Goals Collect great data from sensing platforms Get data from platforms to processing teams Get raw and processed data to data centers End users get data and use it Update or create new request for sensing Rinse – repeat Data volumes are growing fast! Currently KB to MB to GB to TB / day for sonars. Deep sonar: em302 collects 8GB / day Shallow sonar: Reson 7125 collects 2 GB/minute.

13 Where are we now Shipping around USB sticks and hard drives – Lots of effort to track physical entities – A fraction are lost or destroyed – Bizarre consequences of government rules Who knows what data has been collected? When will data be available? How long before an issue gets noticed? rsync is the cutting edge of the bathymetry world

14 Where do we want to be? Data sent in priority order (e.g. post huricane survey) Anyone can find out what has been collected and see where it is in the scheme Be able to push a small subset for validation in a style similar to progressive scan images. Shortest time to problem detection. Balance cost and take best advantage of each method of data transfer. Factor in backup with platform data capacity and stability. Do we need the “mail buoy”? Tools that “just know what to do” but allow the user to know too.

15 Issues Peer-to-Peer bad rap from BitTorrent Software technical skill levels generally low Need to think about data going both ways Community stuck on the file metaphor, but really we have streams of observations that range widely in size from bits to MB per observation. e.g. bool (T/F/Unknown) to seismic shot (thousands of hydrophones on streamers) Keeping the channel utilization high Ships not designed for data logging

16 Formats currently in use Raw bits/bytes ASCII – e.g. MGD77, CSV, old sonars ASCII NMEA (proprietary) Containers (e.g. netCDF, HDF5, etc) Proprietary Binary Packet Systems – e.g. (the not so) Generic Sensor Format (GSF) ProtoBuf, ASN.1, JSON/BSON, XML, AIS (w/ English language packet definitions) Streaming video / audio Email etc.

17 Video For some applications, you might want realtime HD video if Channel A is available, reduced quality if Channel B, and otherwise 1 frame per X time.

18 Do we need “mail buoys?”

19 Thanks for listening! http://schwehr.org Want to know more? http://tinyurl.com/da-paper QR Code:

20 QR Code


Download ppt "Delay Tolerant Data Collections Bathymetry in Google Earth dtnrg 2012 Kurt Schwehr Google / UNH CCOM"

Similar presentations


Ads by Google